The human Philadelphia (Ph) chromosome-positive leukemias, including chronic myeloid leukemia (CML) and B-cell acute lymphoblastic leukemia, are among the most common hematological malignancies, and current therapy for these diseases is inadequate. Expression of the product of the t(9;22) Ph chromosome, the BCR!ABL fusion gene, in the hematopoietic system of mice by generation of transgenic mice or through retroviral transduction and transplantation of bone marrow has demonstrated that BCR/ABL is a leukemia specific oncogene and the direct cause of CML. The long term objective of this application is a more complete molecular and genetic understanding of the pathophysiology of human Ph-positive leukemias, particularly the myeloproliferative disease CML. These goals will be accomplished by the use of a retroviral bone marrow infection/transplantation mouse model system that accurately and quantitatively models both human CML and Ph-positive B-lymphoid leukemia, and will have two Specific Aims. In the first Aim, the signaling pathways important for leukemogenesis by BCR/ABL will be identified by testing BCR/ABL mutants, by further analysis of the requirement for direct binding of the Grb2 adapter protein by the Bcr/Abl fusion protein, and through the use of mice with germline mutations in signaling molecules. In the second Aim, the bone marrow target cells that initiate the CML-like disease and B-lymphoid leukemia induced by BCR/ABL will be characterized and isolated by physical and immunological methods. These investigations will add important new information to our understanding of these leukemias, that would be difficult if not impossible to obtain from studies in vitro, in cultured cells, or in primary human CML cells. This knowledge will be valuable for improving the diagnosis and treatment of the Ph-positive leukemias.